ID3 algorithm for decision tree analysis and optimization
Fan Tai-hua · Jisuanji gongcheng yu sheji · 2012
First,ID3 algorithm's basic principles and major shortcomings,and advantages and disadvantages of several existing improved algorithms are simply analyzed by this paper.Then for ID3 algorithm the main drawback that tends to select the attribute which has more values,which has been significantly improved by using the rough set theory and mathematical know-ledge points.Theoretical analysis and experimental results show that the improved algorithm,to a certain extent,not only can well solve the multi-valued bias problem of ID3 algorithm and greatly simplify the computational process,obviously improve the algorithm's classification accuracy and implementation efficiency.